The present invention relates to a charge-pump circuit outputting converted voltage in increments of less than a power supply voltage Vdd, and a control method thereof, particularly a control method of a charge-pump circuit capable of normal charge-pumping operation removing influence of a parasitic diode associated with a charge transfer device.
The charge-pump circuit devised by J. F. Dickson generates higher voltage than power supply voltage Vdd of an LSI chip by connecting plural stages of the pumping packet in series to boost the voltage. For example, it is used for generating voltage for program/erase of flash memories.
However, the conventional charge-pump circuit carries out voltage conversion in increments of the power source voltage Vdd, and a circuit capable of carrying out voltage conversion in increments of less than the power supply voltage Vdd had not yet been proposed. Efficiency of the power supply circuit can be significantly improved, if a voltage boosting in increments of e.g. 0.5 Vdd is realized.
So, the inventor has already proposed a charge-pump circuit capable of carrying out voltage conversion in increments of less than the power supply voltage Vdd and improving efficiency xcex7 of the circuit (U.S. patent application Ser. No. 09/732,944 filed on Dec. 8, 2000).
The outline thereof will be described below. FIG. 9 and FIG. 10 are circuit diagrams showing a circuitry and operation of +0.5 Vdd voltage boosting charge-pump circuit. The charge-pump circuit generates a boosted voltage of 1.5 Vdd from an input voltage Vdd.
Diodes D1 and D2 (charge transfer devices) are connected in series. The power supply voltage Vdd is provided as an input voltage Vin to an anode of the diode D1. Each of the diodes D1 and D2 can be realized with a MOS transistor with its gate and drain connected together. Switches S1, S2 and S3 connect two capacitors 1 and 2 to a connecting point between the diodes D1 and D2, alternating between in parallel and in series.
The switches S1, S2 and S3 can be implemented with MOS transistors. A clock driver 3 provides the capacitors 1 and 2 with clock CLK. An output of the clock driver 3 is connected to one of the terminals of the capacitor 2. A power supply voltage to the clock driver 3 is Vdd, and amplitude of the clock is Vdd. The clock driver can be implemented with two stages of CMOS inverters for example.
The boosted voltage outputted from the diode D2 is supplied to a load 4. The output node of the diode D2 has a capacitance CL.
Next, the operation of the charge-pump circuit will be explained referring to the FIGS. 9, 10 and 11. FIG. 11 is an operating waveform chart of the charge-pump circuit. For the sake of simplicity, a voltage drop VF of the diode D1 and D2 is assumed 0 V, and the capacitances of the condensers C1 and C2 are assumed equal.
When input clock CLK of the clock driver 3 is at low level (CLK=Low), and assuming that S1 is OFF, S2 is ON and S3 is OFF, two capacitors 1 and 2 are connected in series to a connecting node (pumping node) between the diodes D1 and D2, as shown in FIG. 9. Thus the voltage at the connecting node V1 is Vdd and each of the capacitors 1 and 2 are charged to 0.5 Vdd respectively.
At this time, a current lout from the power supply Vdd flows through the diode D1, and into the capacitors 1 and 2. The same current lout flows into the clock driver 3.
Next, as shown in FIG. 10, the input clock of the clock driver 3 changes to high (CLK=High), and the switches turn to S1=ON, S2=OFF and S3=ON. Thus the two capacitors 1 and 2 are connected to the connecting node of the diodes D1 and D2 in parallel. Since the voltage of each of the capacitors 1 and 2 is Vdd/2 and the output voltage of the clock driver 3 is Vdd, the voltage V1 at the connecting node (pumping node) between the diode D1 and D2 is boosted to 1.5 Vdd.
At this time, a current which flows from the two capacitors 1 and 2 to the diode D2 is 2 Iout, and the same amount of current 2 Iout flows out from the clock driver 3.
Followings stand in a steady state, assuming that the output current from the diode D2 is steady at 2 Iout, and taking average current over time for each current.
1) Vout=1.5 Vdd where the power supply voltage of the driver is Vdd.
2) Input current=0.5 Iout
3) Current from the power supply Vdd of the clock driver=Iout.
An important point of this embodiment of the charge pump circuit is that the boosting in increments of 0.5 Vdd is done by repeating charging the capacitors 1 and 2 connected in series and discharging them connected in parallel. It is also important that the input current Iin=Iout from the power supply Vdd is a xc2xd of the output current 2 Iout. Because of this, the theoretical efficiency xcex7 of the circuit can be 100% and there is no loss in boosting to 1.5 Vdd, when no regulation is made on the output voltage.
The input current is a sum of 2 Iout while CLK=H and Iout while CLK=L. Thus,
the efficiency xcex7 of the charge-pump circuit
=output power/input power
=(1+0.5)xc2x7Vddxc2x7Iout/Vddxc2x7(1+0.5)xc2x7Iout
=1=100%
Therefore, the charge-pump circuit described above can be regarded in effect as a 0.5 stage charge-pump circuit. And the efficiency xcex7 of the circuit can be made 100%. The voltage of 0.5 Vdd can be generated by other methods, for example by a resistance voltage divider. However, it can not realize the efficiency xcex7 of 100%, being a certain power loss inherent in it.
On the other hand, the power loss can be made 0% theoretically with the charge-pump circuit proposed by the inventor, since the connection of the capacitors are alternated between series and parallel in accordance with the level of the clock CLK.
If the two capacitors 1 and 2 are kept in series (S1=OFF, S2=ON, S3=OFF) during the operation regardless of the level of the clock CLK, then the circuit operates in the same way as the conventional charge-pump resulting in Vout=2 Vdd. In this case, a switch control circuit (not shown) provides the switches S1, S2 and S3 with the switch control signal whether to keep the capacitors 1 and 2 always in series or to alternate between series and parallel in accordance with the level of the clock CLK.
That is to say, the charge-pump circuit of the embodiment can generate either 1.5 Vdd or 2 Vdd. In other words, it can be switched between 0.5 stage and 1 stage charge-pump.
However, a detailed examination by the inventor had proved that a malfunction occurs in the transition from the status of the FIG. 9 to the status of the FIG. 10 as well as in the reverse transition, if the timing to switch the S1, S3 and S3 is not aligned properly.
For example, in the transition from the status of the FIG. 9 to the status of the FIG. 10, the voltage V1 at the connection between the diodes D1 and D2 becomes 2 Vdd, if the clock CLK is changed from L level to H level while the switch S2 is ON.
Also, in the transition from the status of the FIG. 10 to the status of the FIG. 9, the voltage V1 at the connection between the diodes D1 and D2 becomes 0.5 Vdd, if the clock CLK is changed from H level to L level while the switches S1 and S3 are ON.
An object of the invention is to prevent the malfunctioning of the charge-pump circuit which boosts a voltage in increments of less than the power supply voltage Vdd.
A control method of the charge-pump circuit of this invention is to align the timing of the change in the clock CLK and the switching of the S1, S2 and S3. The control is done in steps {circle around (1)}-{circle around (7)} described below.
{circle around (1)} The switch S2 (the first switching means) is turned ON to connect the capacitors 1 and 2 in series while the clock CLK is at L level.
{circle around (2)} The switch S2 is turned OFF.
{circle around (3)} The clock CLK is changed to H level.
{circle around (4)} The switches S1 and S3 (the second switching means) are turned ON to connect the capacitors 1 and 2 in parallel.
{circle around (5)} The switches S1 and S3 are turned OFF.
{circle around (6)} The clock CLK is changed to L level.
{circle around (7)} Repeat the steps {circle around (1)} through {circle around (6)}.